Traditional models based on historical load patterns and gradual growth no longer reflect conditions in regions where a single facility can add hundreds of megawatts of demand within months. Modern, interoperable architectures that support real-time data exchange give operators a live view of grid performance and the flexibility to adjust as conditions and power requirements evolve across large-scale digital infrastructure. Data centers vary widely in size, redundancy requirements, and grid connection type, as shown below. In some regions, proposed data centers would require more power than the utility’s existing customer base consumes in total.
One idea floated is a requirement for coordination and ride-through capabilities so that large loads don’t inadvertently destabilize the grid. For instance, FERC Order 2023 (issued in mid-2023) requires transmission providers to implement stricter timelines and a first-ready, first-served cluster study process for generation interconnection. After years of flat electricity demand, many regions are now forecasting substantial growth driven largely by these digital loads, and they’re scrambling to upgrade infrastructure accordingly.
- We design and manage beneficial electrification programs for buildings, vehicles, and industry.
- Instead of depending exclusively on large, centralized clouds, companies are also building regional data centers, colocation footprints, and specialized on-prem deployments optimized for AI inference and emerging agent workloads.
- State Sen. Kevin Hertel, a Democrat from St. Clair Shores outside Detroit, co-authored the data center bills and opposed proposals that would have added some consumer protections and required data centers to build their own clean energy.
- Most utilities already have strong, coordinated storm response and recovery capabilities and should continue to build on that foundation through faster mobilization, better coordination and scaled logistics.
- In 2026, utilities will continue to shift from planning to execution.
- DOE has anticipated this growing demand trend — it reflects robust industrial investments in America and national leadership on technology innovation.
OpenAI CEO Sam Altman, in a letter to the White House’s Office of Science and Technology Policy, recommended that the United States add 100 gigawatts of energy production capacity a year to stay competitive in the artificial intelligence (AI) race. PJM, the country’s largest grid operator, is meanwhile weighing new rules that would require incoming data centers to curb their impact on overall capacity costs. In response, some operators are pledging to shift their electricity use to off-peak hours to ease grid pressures. Peak load growth in the United States is expected to increase by 166 gigawatts over the next five years, according to Grid Strategies — over four times higher than the 2023 estimate of 38 gigawatts and over double the 2024 estimate of 64 gigawatts.
- One idea floated is a requirement for coordination and ride-through capabilities so that large loads don’t inadvertently destabilize the grid.
- Doing so will require new approaches, new thinking and new partnerships.
- These systems provide local electricity to critical loads, reduce peak demand, provide essential reliability services and bring new capacity online faster than traditional projects constrained by interconnection timelines.
- Utilities are starting to recognize that achieving this reliability requires renouncing allegiances to net-zero goals and accepting “the inherent advantages of stock-versus-flow energies.”
- Total power generation for renewables is projected to grow 22% each year until 2030, meeting nearly half of the anticipated growth of data center electricity demand.”
- With AI-driven computing expanding rapidly, the scale and speed of data center development are increasing the risk of interconnection delays, infrastructure bottlenecks, and costly stranded investments if planning is not carefully coordinated.
How to Add Us to Google News
Faced with grid interconnection delays of five years or more, hyperscalers will increasingly build dedicated generation facilities alongside their data centers. With several other retired plants in regions of high data center demand, the economics strongly favor additional restarts, contingent on regulatory approval and safety assessments. If SMR technology matures as proponents hope, it could provide a scalable solution for powering individual data center campuses with dedicated clean energy.
Notable H2 2025 North American Capital Markets Activity
In Texas, for instance, large users interconnecting after 2025 will be required to pay retail transmission charges based on peak demand. Electric companies are expected to make massive investments to modernize the grid to address growing electricity demand. Key inflection points will likely include the repeal or phaseout of certain clean energy tax credits, evolving tariffs, new foreign entity of concern–related procurement requirements, and the integration of AI into core operations.
Key transactions between November 2025 and May 2026:
The Task Force plans to expand its work on affordability and ratepayer protections this year. The Pennsylvania Public Utility Commission (PUC) has been developing a model large-load tariff that will serve as a statewide framework to help ensure that large-load customers can connect to the grid quickly and responsibly, while also supporting long-term reliability. Some states are shifting from relying on utilities to propose their own large-load tariffs toward a more top-down, state-driven approach to developing tariffs and service rules. It marks a foundational shift in how states perceive the obligations of large power users like data centers. Each state is addressing the challenge through its own lens, often driven by local political priorities or the concerns of a single regulator or legislator.
Utilities use the peak load forecasts to justify investments in power plants and grid infrastructure, which is driving up utility bills in some regions, including the PJM Interconnection, the country’s biggest energy market. The Department https://214rentals.com/garage-construction-in-edmonton-basic-requirements-and-advantages-of-contacting-professionals.html continues to develop advanced technologies and leverage its resources to meet rising electricity demand in the United States while maintaining a reliable, affordable, and secure national energy system. DOE has anticipated this growing demand trend — it reflects robust industrial investments in America and national leadership on technology innovation.
Utilities and large energy users are increasingly deploying onsite power—including microgrids—to help manage the strain. In addition, utilities are expanding networks of high-voltage transmission lines to maintain system balancing and to ensure reliable electric service. Expected electricity demand growth is spurring expansion in generating capacity and electricity storage. This growth contrasts with the trend of relatively flat electricity demand between the mid-2000s and early 2020s. Foley can also assist companies involved in mining, processing, and disposal to navigate the various state and federal regulations, international agreements, and national security considerations. The group counsels clients regarding key governmental processes and areas of law based upon decades of collective legal or governmental experience.
- As AI workloads scale from pilots to production, experts say 2026 will test the limits of data center energy, operations, and sustainability.
- The EIA flagged that residential electricity prices are expected to increase by 5% in 2026, prompting regulators to favor firm, dispatchable generation.
- Increased U.S. load growth from data centers and manufacturing is expected to put stress on grids across the country, which could lead to blackouts if severe weather hits.
- “Connection lead times of one to two years, demands for highly reliable power, and requests for power from new, non-emitting generation sources can create local and regional electric supply challenges,” the report found.
- On the demand side, advances in chip design, cooling systems, model optimization, and compute efficiency could lower power intensity per unit of AI workload, reducing expected electricity demand growth relative to current projections.
- Applications will require proof-of-site control and financial commitment to build necessary infrastructure, in addition to a minimum $100,000 fee for initial transmission screening.
Large, concentrated loads are emerging faster than traditional planning cycles https://eurodialogue.org/energy-security/Energy-Are-There-Limits-To-Growth were designed to accommodate, challenging existing approaches to forecasting, infrastructure development, and system reliability. By bringing the data centers and utility companies together to learn from each other and develop a path for a better future.” These often require dual feeds, with one as primary and another for failover. If your team is dealing with transformer upgrades, you may want to revisit our guide on CTs and PTs to ensure the meter setup is scalable and accurate. Unlike residential or commercial customers whose loads vary based on time of day or season, data centers run nearly constant 24/7 loads.
This shift is creating sites that are equivalent to “small cities” with highly volatile load profiles driven by AI, data and cloud computing. Data centers, transportation fleets, buildings and industrial operations are all rapidly transitioning from molecules (fossil-fuel-based systems) to electrons (electric power), creating simultaneous, unprecedented load growth. With approximately 60% of the energy consumed by data centres today coming from fossil fuels, companies will be forced to choose between AI capabilities and environmental commitments – a tension that will drive innovation at an unprecedented pace.” AI training and consumption will require more data centres, and the locations and energy sources of those data centres will come under intense scrutiny. With this transition, metrics for success are shifting from sustainability and conventional efficiency toward revenue generation.
The major goal of a capacity market auction is to reward power suppliers that invest in capacity to meet future peak load requirements to ensure reliable operation of the grid. Workloads are increasingly segmented based on their individual requirements, from processing speed to privacy laws – and it is all in incredibly high demand. From on-site cogeneration and load balancing to demand-response participation and remote-curtailment capabilities, utilities and regulators are increasingly embedding flexibility requirements into contracts and permitting. The deferred fuel cost charge is a mandatory charge required by the Virginia SCC, allowing utilities to recover previously incurred, unanticipated fuel charges in future billing periods. It requires new data center customers in Ohio to pay for a portion of their energy requests, even if they use less, to cover the cost of the infrastructure required to bring electricity to those facilities. On the demand side, advances in chip design, cooling systems, model optimization, and compute efficiency could lower power intensity per unit of AI workload, reducing expected electricity demand growth relative to current projections.